13. the Northern Spice Islands in Prehistory, from 40,000 Years Ago To
Total Page：16
File Type：pdf, Size：1020Kb
13 The Northern Spice Islands in prehistory, from 40,000 years ago to the recent past Peter Bellwood The previous 12 chapters have examined the results of our 1990s archaeological project in the Northern Moluccas from the perspectives of chronology, artefact sequences, animal remains, and human remains. The general goal has been to locate the Northern Moluccas within the Island Southeast Asian record of humanity since approximately 40,000 years ago. From start to finish, a number of significant questions have risen to the surface: a. What kind of hominins first settled the Oceanic (non-landbridged) islands of the Northern Moluccas, and when? b. Were these islands settled permanently by forager populations before the commencement of food production, or did these populations only utilise the smaller islands, such as Gebe, intermittently? c. Were marsupials translocated from the New Guinea region to the Northern Moluccas by human intention, or did these animals disperse naturally? d. What evidence exists within our data for interaction between the island societies of Southeast Asia (including the Moluccas) and western Oceania during the Late Pleistocene and Early Holocene, before the Neolithic? e. How do sites such as Uattamdi fit within the picture for Neolithic dispersal around 3300 years ago through Island Southeast Asia and into Oceania? Were there links with the Lapita archaeological culture in Island Melanesia and with the initial settlement of the Mariana Islands? f. What have been the traceable interactions between Austronesian-speaking and Papuan- speaking societies in the Northern Moluccas/New Guinea region during the past 3500 years, and especially the past 2000 years since the spice trade linking western and eastern Eurasia commenced? When did pottery making spread into the Papuan-speaking societies of Morotai and northern Halmahera? Furthermore, were the ‘Spice Islands’ actual sources of exported spices such as cloves around 2000 years ago, as a lot of rather loose historical inference inclines to inform us? First, let us examine chronology. In terms of standard archaeological terminology, four overall phases (Northern Moluccas Phases 1 to 4) can be recognised in the prehistory of the Northern Moluccas. The earliest record of a potential human presence comes from Golo Cave on Gebe, with a basal date of 36,350–35,001 cal. BP (Wk 4629). Beyond this followed many millennia of Palaeolithic occupation, possibly intermittent, with bone tools and edge-ground technology 212 The Spice Islands in Prehistory for shell (but not stone) adzes appearing towards the end of this phase; indeed, during the Early Holocene rather than the Late Pleistocene in terms of current chronology. We can call this first phase Northern Moluccas Phase 1, let us say from 40,000 to 3500/3000 years ago, with an intensification of cultural activity during the Early and Middle Holocene. A major change occurred associated with the oldest date of 3342 to 2971 cal. BP (ANU 7776) from Uattamdi, which marked the appearance of red-slipped pottery and the oldest shell ornaments found so far in the region. This second, Neolithic, phase apparently continued in Uattamdi until about 2000 BP, and is here termed Northern Moluccas Phase 2. The first appearances of cupreous and iron artefacts and glass beads ushered in a third, metal-using, phase (Northern Moluccas Phase 3), again at Uattamdi, by C14 date ANU 9322, 2136 to 1782 cal. BP. This Early Metal Phase also saw the first recorded appearances in the region of jar burial (Uattamdi), supine orientation for an inhumation burial (Golo Cave), and a spread of earthenware potting technology into nearby previously aceramic communities on Morotai and northern Halmahera (Tanjung Pinang, Aru Manara, Gorua). The Early Metal Phase continued onwards into the historical period (Northern Moluccas Phase 4), commencing in these islands in terms of direct written records with the arrival of the Portuguese in 1511 (the manuscript of Tomé Pires was written between 1512 and 1515; Cortesão 1944). One could perhaps also recognise a separate phase of importation of glazed ceramics from Chinese and other mainland sources during the early second millennium CE, but little such material was found during our excavations. To summarise these four phases, therefore, we have the following: Northern Moluccas Phase 1: from 40,000 to 3500/3000 BP, flaked stone and shell from the beginning, but with marsupials, bone tools, and edge-ground technology for shell adzes only appearing definitely in Holocene but pre-Neolithic contexts. Northern Moluccas Phase 2: 3500/3000 to 2000 BP, Neolithic red-slipped pottery, shell ornaments, and pig and dog bones (no data on human burials). Northern Moluccas Phase 3: 2000 to 500 BP, Early Metal Phase cupreous and iron artefacts, glass beads, jar burial (Uattamdi), skull burial (Tanjung Pinang), extended burial with cranial modification (Golo), spread of pottery into Morotai and Halmahera (see, on this, Ono et al. 2017, accepted), extirpation of marsupials, spice trade(?). Northern Moluccas Phase 4: 500 BP to recent, European contact, Asian trade ceramics. With this standard archaeological periodisation in the background, a number of important issues arise. Some can be debated fruitfully, others remain obscure, and many have been discussed in more detail in my newest book-length survey of the region (Bellwood 2017). Some outstanding questions What kind of hominin first settled the Oceanic (non-landbridged) islands of the Northern Moluccas, and when? With archaic hominins attested in Java and Flores by Middle Pleistocene times, and possibly also Sulawesi and Luzon by the Late Pleistocene, one might ask if they also reached the entirely sea-girt islands of eastern Wallacea, even allowing for Pleistocene fluctuations in sea level (Hope 2005; O’Connor et al. 2017). Unfortunately, Pleistocene lithic technology in Island Southeast Asia reveals nothing about the hominin status of the makers, so we can interpret very little of a taxonomic nature from the lithic record in Golo Cave at >36,000 years ago. Neither terra australis 50 13. The Northern Spice Islands in prehistory, from 40,000 years ago to the recent past 213 in Golo nor in other contemporary sites in Island Southeast Asia are there Pleistocene blade, microblade, or backed tools that might give the species game away and hint whether we are dealing with Homo erectus, dwarfed hominins related to Homo floresiensis, or Homo sapiens. The Pleistocene paintings in caves in Sulawesi and Borneo, the widespread terminal Pleistocene bone tools, and the eventual settlement of Australia and New Guinea by Homo sapiens at about 50,000 years ago or more would suggest that the first traceable Moluccans at 36,000 years ago were indeed Homo sapiens. But we cannot be sure, and no human remains in Moluccan sites are old enough to be of assistance. However, as Katherine Szabó has pointed out (in Szabó et al. 2007), the presence of modified shell items in the basal layers of Golo Cave suggests a potential presence of behaviourally modern Homo sapiens there, as opposed to an archaic hominin. But, without dated cave art and skeletons, a degree of uncertainty remains. We might also ask just when humans, modern or otherwise, might first have settled in the Northern Moluccas, noting that it is just as likely that Halmahera was settled initially from western New Guinea, via Gebe, as from any other source region to its west (Birdsell 1977; Kealy et al. 2017). New Guinea was certainly settled by 50,000 years ago (Summerhayes et al. 2010), but this need not mean that every small island in its vicinity was settled also at this time. Indeed, Golo and Wetef were both dug to bedrock, with a maximum date of only 36,000 years ago, so we cannot argue that older material lies unexcavated beneath. Gebe is such a small island that if people really did arrive long before 36,000 years ago, we might ask where they lived without leaving traces in any of the excavated caves and rockshelters on the island. All in all, human arrival at 36,000 years ago on Gebe is the best we can do at the moment, and a claim for anything older would be unwarranted. Were these islands settled permanently by forager populations before the commencement of food production, or did they utilise the smaller islands, such as Gebe, only intermittently? Only Golo and Wetef on Gebe Island have sufficient time depth for analysis of this question, given the absence of Pleistocene occupation within Tanjung Pinang on Morotai, and the very limited evidence for it from Daeo 2. Indeed, there is a possibility that humans did not occupy Morotai at all until after the Last Glacial Maximum, although this is purely a negative observation that could easily be overturned by future discoveries. The Wetef dates (Table 2.5) from Gebe certainly indicate a consistent human presence during the Holocene, as they do in Golo. But Wetef had rather limited Pleistocene occupation, and Golo has some obvious problems with the distribution of C14 dates. In Table 2.2 for Golo, I have shaded the dates that appear to give a coherent sequence, and I regard all the ‘rogue’ marine shell dates as perhaps reflecting disturbance due to the frequent caching of shell adzes in the soft dry floor of the cave. The three direct dates on the shell adzes themselves cannot be taken seriously as indicators of stratigraphic date, as discussed in Chapter 8. If the apparent rogue dates are ignored, then the Golo sequence between 30,000 and 13,000 years ago could suggest either extremely slow deposition, or even an occupation hiatus. At this point, the Wetef date series might come to the rescue, since they do indicate occupation in the lower levels dated to c. 23,000 BP (Table 2.5).